Adv Top Sig Proc Proteomics
Adv Top Sig Proc Proteomics EE 6363
Popular in Course
Popular in Electrical Engineering
This 15 page Class Notes was uploaded by Valentine White on Thursday October 29, 2015. The Class Notes belongs to EE 6363 at University of Texas at San Antonio taught by Staff in Fall. Since its upload, it has received 21 views. For similar materials see /class/231450/ee-6363-university-of-texas-at-san-antonio in Electrical Engineering at University of Texas at San Antonio.
Reviews for Adv Top Sig Proc Proteomics
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 10/29/15
Aerial Images Directional Denoising by Splitting Signals Fatma T Arslan Andrew K Chan and Artyom M Grigoryan Department of Electrical Engineering University of Texas at San Antonio Outline 0 Introduction 0 Tensor Representation of Images 0 Wavelet Denoising Methods 0 Directional Denoising of Aerial Images by Tensor Trans form 0 Conclusions IEEEREGS Aerial Images Directional DenOISIng by Splitting Signals Introduction The oceonagraphic aerial images are commonly used for study ing of ocean current flow seabed structures rock locations sediment formation etc Usually these aerial images are cap tured with wave clutters The clutters are classified into two types 0 ripple wave long waves o spark wave short waves Waves are added to the test images These clutters are directionally located Aerial Images Directional Denoising by Splitting Signals IEEEREGS Introduction Tensor Transform are used to denoise these images By Ten sor Transform images are represented in the form of 1 D independent splitting signals The splitting signals show certain directional effects in fourier domain In this study we benefit from these features of splitting signals Aerial Images Directional Denoising by Splitting Signals IEEEREGS 2 a A New Look on Image and 2 D DFT o Analyze the tensor and paired representations of images in forms of sets of 1 D independent splitting signals o Tensor Representation transfers image to 3N2 splitting signals of length N o Seperately process all or only a few splitting signals and then calculate and compose the 2 D DFT of the processed image by means of new 1 D DFTs of the processed splitting signals Aerial Images Directional Denoising by Splitting Signals IEEEREGS Tensor Representation LLD DPT O a o V was l ooo x f IDDNM 3 3 ogg Qal erDFT DC fT1fT2vf1393fT4 T17T27T37T4 Diagram of 3 x 3 point sequence f transformation into four signals fT1fT2 fT3fT4 whose 1 D DFT defines the 2 D DFT of f Aerial Images Directional Denoising by Splitting Signals IEEEREGS Tensor Representation The irreducible covering a T of XNyN is defined by the following cyclic groups with generators ps Tm 0 0 195 27975 N 1mm 1s 1 For instance the covering of the 3 x 3 set X is defined by 0 T10T11T12T01 as Show C C C C O O O O C O C O O O O o o o o o o o o o o o o o o o C C C C O O C O O C O O C C C X33 T10 T11 T1 2 To1 Aerial Images Directional Denoising by Splitting Signals IEEEREGS 4 a Tensor Representation The covering a T1975 reveals the 2 D DFT The following property holds for the Fourier transform N l where fpys f Z fn7m7 t O Vpst and prsyt nm np I ms t mod N Aerial Images Directional Denoising by Splitting Signals IEEEREGS Tensor Representation 130 125 120 115 1 O splitting signal fT 41 105 100 a 05 04 03 DFT0150 T 02 01 O m 1 H 1 m 15 200 400 C d a Original image b Splitting signal fTM c The 1 D DFT of the splitting signal d Arrangement of values of the 1 D DFT in the 2 D DFT of the image IEEEREGS Aerial Images Directional Denorsrng by Splitting Signals Wavelet Denoising Methods The algorithm used in this paper to denoise 1 D signal is described as follows 0 Perform the forward wavelet transform with j level 0 Discard certain coefficients that has noise a Reconstruct the denoised signal by the inverse wavelet transform The coefficients that are to be discarded are all high fre quency coefficients at j 1 levels Aerial Images Directional Denoising by Splitting Signals IEEEREGS Directional Denoising by Tensor Transform a We propose a novel method combining Tensor Transform by wavelet transform and denoise these images a To remove the long waves subband filtering technique is used 2 D subband filtering is reduced to 1 D filtering by using Tensor Transform To remove the short waves SMEME filter is used Tensor Transform is used as pre step of SEME filter Aerial Images Directional Denoising by Splitting Signals IEEEREGS Directional Denoising by Tensor Transform a b a Aerra Image 1 b Aena Image 2 EEEREGS Aena Images Drrectrona Denorsrng by Sphttrngsrgna s 67a Directional Denoising by Tensor Transform Modelling long waves and short waves Long waves are modeled as 27r Cripplexa90 A Sm Ta gt 3 4 T a 04 g or controls the initial period of long waves B controls the variation of the ripple frequencies A amplitude of the wave yo location at which the long wave occurs The cripple function has higher frequency near the locations where 1 O and lower frequency far from there Aerial Images Directional Denoising by Splitting Signals IEEEREGS 6 b Directional Denoising by Tensor Transform short waves are modeled as c 33 A spark y 1 1 07y 5 2 2 MW 2 a ago y 92 0340 location of the short wave n affects the shape of waves a b and D0 determine the width and length of each wave A amplitude of the short wave peak Aerial Images Directional Denoising by Splitting Signals IEEEREG5 6 c
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'